Anaheim’s Case For At-Large Districts: Lies, Damned Lies, and Statistics

ballotInitiativeTomorrow Judge Franz Miller is set to decide whether the California Voting Rights lawsuit against the City of Anaheim will move forward. Miller had delayed his decision to allow the city to let its Community Advisory Committee determine whether there is a need for council districts to address allegations that the current at-large system disproportionately dilutes the ability of a historically underrepresented minority (Latinos) to achieve representation in city government.

Rather than accept the advice of their own appointed committee, stacked in the favor of the current majority, to place a Charter amendment recommendation before the City establishing council districts that would elect their own representation, the council majority decided to continue at-large voting for representatives that must live within specific district boundaries.

Lies, Damned Lies, and Statistics

The council majority has apparently based its decision on a flawed report from their hired demographer, Peter Morrison. They claim that the Community Advisory Committee would have likely made the same determination if it had the benefit of the analysis included in the demographers report. The problem here is that they rely on a report that relies on flawed analysis which concludes that a correlation of events establishes a causal relationship; otherwise known as a logical fallacy. Two events can consistently correlate with each other but not have any causal relationship. In the case of Morrison’s analysis, he concludes that since a Latino candidate has come in third place in the last two council elections, that it is only a matter of time until Latinos will receive proportionate representation on the council. Morrison also concludes that since two Latino candidates were elected in 2002 and 2006, that this if proof that their representation is not diminished. Morrison conveniently leaves out the fact that the two Latino’s elected were retired firefighters elected in the first election after the 9/11 terrorist attacks, when support for public safety was at a high point in all communities. This would be called an outlying event and most statisticians would would exclude from their calculations.

In simple terms, that should make the validity of Morrison’s conclusions readily apparent lets consider a similarly invalid correlation proposed by Chris Blattman on his blog.

IE-Market Share and Murder Rate

The correlation above proposes that there is a direct connection between the market share for Internet Explorer and the number of U.S. murders over a six year period of time. Hence, Internet Explorer market share must have a causal relationship related to the incidents of murder in the U.S.

Deeper Analysis of Morrison’s Report

We cannot do much better a job of detailed analysis, and debunking, of Morrison’s report than Greg Diamond has over at the Orange Juice Blog. So we refer you to his analysis (here, here, and here) for further confirmation of the fact that the City council majority is basing it’s entire case on flawed statistical analysis, that flakes apart with a simple scratch to the surface.

We wait with breathless anticipation to see if the Anaheim City Council majority’s flawed reasoning holds any sway with Judge Miller tomorrow. Something in our logical minds tells us that their hopes are based upon a logical fallacy.

  2 comments for “Anaheim’s Case For At-Large Districts: Lies, Damned Lies, and Statistics

  1. cook
    July 8, 2013 at 10:35 am

    When the school board guy and ACLU lose tomorrow on a weak argument, are they going to have to pay the court costs and the cities lawyer costs?

  2. Greg Diamond
    July 8, 2013 at 1:15 pm

    I’m not sure, cook. You’d have to check the rules of civil procedure for the District of Fantasyland.

Comments are closed.